Prediction of Inlet Distortion Transfer Through the Blade Rows in a Transonic Axial Compressor

• Main Author: Ryman, John Franklin
• Other Authors: Mechanical Engineering
• Format: Others
• Published: Virginia Tech 2014
• Subjects: Total pressure distortion; Inlet distortion; Distortion transfer; High Cycle Fatigue
• Online Access: http://hdl.handle.net/10919/43207; http://scholar.lib.vt.edu/theses/available/etd-06122003-092045/

Inlet total pressure non-uniformities in axial flow fans and compressors can contribute to the loss of component structural integrity through high cycle fatigue (HCF) induced by the excitation of blade vibratory modes. As previous research has shown total pressure distortion to be the dominant HCF driver in aero engines [Manwaring et al, 1997], an understanding of its transfer through, and impact on, subsequent turbomachine stages and engine components is an important topic for assessment. Since current modeling techniques allow for total pressure distortion magnitudes to be directly related to blade vibratory response, the prediction of downstream distortion patterns from an upstream measurement would allow for the inference of the vibratory response of downstream blade rows to an inlet total pressure distortion. Nonlinear Volterra theory can be used to model any periodic nonlinear system as an infinite sum of multidimensional convolution integrals. A semi-empirical model has been developed using this theory by assuming that a distortion waveform is a periodic signal that is being presented to a nonlinear system, the compressor being the system. The use of Volterra theory in nonlinear system modeling relies on the proper identification of the Volterra kernels, which make up the transfer function that defines the systemâ s impulse response characteristics. Once the kernels of a system are properly identified, the systemâ s response can be calculated for any arbitrary input. This model extracts these kernels from upstream and downstream total pressure distortion measurements of a transonic rotor of modern design. The resulting transfer function is then applied to predict distortion transfer at new operating points on the same rotor and compared with the measured data. The judicious choice of distortion measurement data allows predictions of the downstream distortion content based on a measured non-uniform inlet flow at conditions different from those at which the transfer function was derived. This allows for the determination of downstream total pressure distortion that has the potential to excite blade vibratory modes that could lead to HCF under operating conditions other than those at which the data was taken, such as varying inlet distortion patterns, mass flow settings, rotational speeds, and inlet geometry. This report presents the creation of a Volterra model in order to predict distortion transfer in axial flow fans and compressors. This model, in three variations, is applied to a variety of distortions and compressor operating conditions as measured in the ADLARF tests at the Compressor Research Facility. Predictions are compared with data from the test and final results are also compared with two previous studies conducted at Virginia Tech using the same experimental data. Using the Volterra model it is shown that, with appropriate limitations, distortion transfer can be predicted for flow conditions different from those used for calibration. The model is considered useful for both performance and HCF investigations. === Master of Science